Converting Modifed Sine Wave to Sine Wave

[owen_a]

Good evening,

I bought my self my first Inverter rated at 500W which has a peak rate of 1000W (1kW) maximum. Since some devices such as my main computer for one takes 600W, It does not like the modified sine wave. I was wondering if there is a way to somehow 'convert' the modified sine wave to a standard sine wave so that most of my equipment will work. I'm talking about converting it as soon as it comes out of the 'plug' output of the inverter. Can this be done? If so, how?

Thanks.

 

[berkeman]

Good evening,

I bought my self my first Inverter rated at 500W which has a peak rate of 1000W (1kW) maximum. Since some devices such as my main computer for one takes 600W, It does not like the modified sine wave. I was wondering if there is a way to somehow 'convert' the modified sine wave to a standard sine wave so that most of my equipment will work. I'm talking about converting it as soon as it comes out of the 'plug' output of the inverter. Can this be done? If so, how?

Thanks.

How "modified" is the sine wave? Modified in what way? Lots of switching harmonic content?

You could potentially use a 1:1 Isolation Transformer to smooth out the sine wave, but it would have to be oversize for the power delivery, because of the extra heating that would be caused by the harmonic content. And you will lose some power in the transformer losses...

 

[jim hardy]

Good evening,

...Since some devices such as my main computer for one takes 600W, It does not like the modified sine wave...

What is the symptom that you see?

Inverters tend to make a flat topped sinewave which older electronics actually do like - it makes for less ripple after rectification.

Newer electronics have power factor correction and harmonic suppression circuitry -
i'd be curious, for my own edification, to hear whether that is source of your "trouble" and how it manifests.

old jim

 

[owen_a]

When applying power to the computer, It acts like it is "pulsing". If I turn the power supply off on the computer, it has a continuous humming noise and the inverter fan spins up. However, when I turn on the switch, it starts pulsing again.

How "modified" is the sine wave? Modified in what way? Lots of switching harmonic content?

I'm not sure how exactly modified it is, but I do have a picture of the specification table from the manual.

See here.

 

[skeptic2]

Could the pulsing be, not because of the modified sine wave, but because you are drawing more than the rated power? Does it still pulse if you put a smaller load on it?

 

[owen_a]

That's what I was thinking. It doesn't pulse when powering a standard 230-240V UK Light bulb, or even my TV and Xbox.

 

[owen_a]

Anyone?

 

[jim hardy]

What you describe sounds to me like this:

Computers have a quite large filter capacitor in their switching power supplies.
When first powered on they draw immense current to charge that capacitor.
That can cause the inverter to think it's connected to a drastic overload, perhaps a short circuit, so it shuts down and retries again hoping somebody has removed the overload in meantime.

You need either a bigger inverter
or an "Inrush Current Limiter"
http://www.mouser.com/Circuit-Protection/Thermistors/Inrush-Current-Limiters/_/N-axfxj/

image courtesy mouser electronics

Without actual measurements that's my best guess.

 

[skeptic2]

If you attach an AC voltmeter across the output with the computer connected and running, is the voltage correct and not pulsing?

Perhaps you also should connect 600W of lightbulbs to determine whether the pulsing is due to the inverter or the computer.

(Why do we say AC voltmeter instead of AV meter?)

 

[jim hardy]

Be aware lightbulbs have inrush currrent about 10X normal draw. Screw them in one at a time.

 

[skeptic2]

Be aware lightbulbs have inrush currrent about 10X normal draw. Screw them in one at a time.

I agree that the cold resistance of a light bulb is about 1/10 that of a hot one but I have always been suspicious of the huge inrush current claim. The worst case would be if the bulb were turned on at or near the peak of the cycle and at that point there would be an initial but very short current surge. The high current would heat the filament very rapidly and as it heats up, its resistance increases and it heats up even faster. I suspect that the current surge would last a small fraction of a cycle and the surge at 10X normal draw could be measured in uS. It's also possible that due to IR drops and line inductance, the 10X figure is never realized.

 

[jim hardy]

Your lamp suggestion is a great one. Just with inverters there are surprises - they behave differently than a wall outlet. They have not much tolerance for even brief overloads and that will fool you into thinking they are defective when they're only mis-applied.


Here's a decent paper that looks at incandescent lamps..
http://www.powerbox.info/ssl/inRushCurrent.asp
he says:

An incandescent bulb typically has an inrush current that is typically 12-15 times the amount of the steady state operating current. Thus for a 60 watt bulb that normally draws about 1/2 Amp steady state current, the inrush current is about 7.5A, when using the high end of the multiplier to be conservative.

The inrush current of a switching power supply all depends on the design of the switching power supply. Usually the inrush current is much higher than that of a 60W incandescent bulb unless special, and more costly, steps are taken to limit the inrush current to lower values. For switching power supplies, an inrush current of 30A is fairly typical for a 120VAC input. “Good” CFLs are engineered to have lower in-rush currents.

You wrote:

I suspect that the current surge would last a small fraction of a cycle and the surge at 10X normal draw could be measured in uS.

A millisecond or two. Depends on the thermal capacity of the filament.
I've measured inrush on 1000 watt lamps. One made a 7.5 kw inverter groan and two tripped it.

The worst case would be if the bulb were turned on at or near the peak of the cycle

Quite true.
Bear in mind a sinewave spends most of its time nearer the peaks than the zero crossings - in fact half its time above 70% (what's sin 45° ?) and a quarter of its time above 90%.


Good luck, guys!

old jim

 

[skeptic2]

Thanks for the link.

The oscilloscope display is very interesting. The time scale appears to be set at 10mS/cm and it appears the switch is closed at about 8mS. The point at which the trace crosses the 4A line appears to be just short of 9mS, perhaps 750uS duration and the point at which it reaches 5A the duration is perhaps 500uS. Nevertheless this is far longer than I expected.